If you hang out in the academic spaces that I do then you have probably heard that Virtual Reality (VR) technologies are going to change the entire identity and landscape of how autistic people receive training and instruction. VR has seemingly been the hyped up technology for years now and you might be wondering to yourself,

Is this VR thing all hype or real hope?

The easy academic answer is, ‘it depends’. The actual response is a bit more complex.

Researchers have been investigating whether VR systems can work for helping autistic people, but their research findings are unclear. This is partially because research procedures vary from study to study. It is also because there is a great deal of variability in how autism plays out from one person to another. So, while there are some promising results, there also seems to be some problems with the field’s  tendency to overgeneralize matters with complicates how results should be interpreted. In contrast to how most work has been considered in this field, an autism researcher in the UK named Sarah Parsons has suggested that we instead consider what technologies are being used, for what audiences, under what contexts, for what purposes, and with what supports. 

So this is exactly what my co-author and I decided to do by conducting a literature review with the goal of uncovering, analyzing, and presenting VR systems have been designed for individuals with autism. More specifically, our review focused on:

1. Comparing how researchers define VR;
2. Examining individual components of VR interventions along the dimensions outlined by Sarah Parson of: (a) which technologies were used, (b) for whom, (c) in which contexts, (d) with what kinds of support, and (e) for what kinds of tasks/objectives; and
3. Examining how VR-based interventions for individuals with autism have been designed.

 But before I get into this, let’s take a step back and ask ourselves…

Why all the hype though?

Well, back in the mid 1990’s researchers in autistic spaces apparently did not get the memo about how destructive VR was to American cinema…and decided to make it a central point of interest.

I am joking of course…kinda. The truth is that there has been a growing interest in using VR as a way of providing interventions for this autistic people since preliminary work was done by Strickland and her colleagues in 1995 to explore the acceptance of using VR headsets with autistic adolescents. Since then, research has continued and many have argued that VR maintains a number of benefits for promoting social and communicative skills for autistic people I myself got pulled into the hype and spent an uncountable number of hours reviewing the literature during my doctorate and assisting in the development of multiple VR systems. It was through this work that I started to notice some problems in the field.

So what are these problems?

Well, let me provide you with some quick bullet points:

1. Autism doesn’t look the same in different autistic people. Autistic differences have been described as a “spectrum,” but, in fact, even that has been criticized as being too simple to describe just how vast the differences are. As the saying goes, ‘If you’ve met one autistic person, you’ve met one autistic person. So what’s the problem? Only a fraction of the research on VR for autistic people describes study participants in enough detail. So we don’t really have a good idea of what is working for whom!
2. It appears that people in the field do not know what Virtual Reality actually is…you might think I am kidding with is, but I assure you I am not. 
3. Authors do not report how VR systems have been designed for from a hardware or a software perspective. Further complicating matters is that this confusion extends well beyond these points and that there are many systems that are described which are objectively not VR.

Woah….WOAH! You might be thinking that I have made some bold claims here (I bolded the word bold. How cool am I?). And you might be saying to yourself

That is a fair point. You know, this entire manuscript actually got started because I told (ranted at) my dissertation chair (co-author) about all of these problems in the field and he told me to prove it. So I did prove it…through a really in-depth analysis of  decades of research in the field across more than 100 published manuscripts over the course of several months. I could bore you with the details of that process, but for now I will skip ahead to the good bits.

Some Findings

• Unfortunately and unsurprisingly, findings suggest that a well-recognized operationalization of what defines VR and how systems are conceived have not yet been established in this field. So even though you might have a clear picture of what VR is in your head, this understanding does not persist in this particular academic space.
• There is no agreement on what a VR technology actually is. Given the first finding this sort of makes sense. If we cannot define ‘virtual reality’ then determining what a VR technology is can also be challenging. We can certainly get into a conversation around hardware vs software and go into the semantics of whether or not there needs to be a physical device assisting in the feelings of telepresence. Unfortunately these conversations do not tend to take place in the literature. Instead, authors vaguely cite some benefits for a loose umbrella term of ‘VR’ while ignoring the vastness of the VR landscape (desktop-based, fully immersive HMD-based, mobile-based, CAVE systems, etc).
• And related to this (and a point that should go without saying) is that the way that VR interventions are designed for individuals with autism (i.e., system architecture and system type) greatly impacts how the system can be designed to promote these benefits and how users can interact with them. However, the field is laden with researchers who are trying to develop solutions for a wide range of target audience demographics and outcomes and the systems being used are often designed in ways that contradict their desired goals. Researchers and developers need to take a more systematic approach to defining and implementing the technology, as well as aligning the technology’s benefits with the needs and abilities of the target audience.
• And lastly, not a single publication clearly stated what specific benefit or characteristic of VR was actually being designed for and used in these systems.

In light of these findings, practitioners and researchers need to better consider their end-users in the ideation, development, and implementation of VR systems. People with autism are not all the same and we need to stop treating them like they are and we need to stop presenting findings from our work in the same light.  Different technologies, with different benefits and characteristics, will only support learning when the system has been designed to specifically consider the needs of the learner and the desired outcomes. VR will not solve the problems in autism research until a more careful approach is taken.

Is there some real hope for the use of VR with autistic people? Sure. But right now the majority of the work is unfocused. Simply put, virtual reality for autism is no silver bullet.

Still want more details?

Such as the boring methods? I don’t blame you. We are some pretty cool people doing some pretty neat work. You can read our full article, Systematic Literature Review of Virtual Reality Intervention Design Patterns for Individuals with Autism Spectrum Disorders or follow our other work. I have related work and information shared on my Research Gate profile and my co-author Matthew Schmidt has a stellar website with more information on his ongoing projects.